Case Study: SteelHouse achieves 100% SLA compliance and stable, high-throughput performance with ScyllaDB

How a Real-Time Switchover from Cassandra to Scylla Saved the Day

SteelHouse, a self‑service advertising platform that processes massive, latency‑sensitive traffic, was struggling to meet strict AdTech SLAs due to inconsistent performance, frequent timeouts, and JVM instability from their Apache Cassandra deployment. To resolve these issues they turned to ScyllaDB, adopting ScyllaDB’s Cassandra‑compatible Scylla clusters to replace their failing Cassandra backend.

ScyllaDB helped SteelHouse perform a live hot‑swap to Scylla, and within about six hours they were running production on Scylla with auto‑tuning, far fewer timeouts, and consistent low latencies. The move eliminated network timeouts, enabled 100% SLA compliance, handled roughly nine billion monthly requests (including peak Black Friday/Cyber Monday traffic), consolidated workloads to ~48 nodes, and cut unscheduled maintenance by 3–4 hours per week—about a 20% productivity gain and lower hardware and operational costs.

SteelHouse

Kevaughn Bullock

Application Architect